Economic geology of offshore gas hydrate accumulations and provinces

Marine and Petroleum Geology 19 (2002) 1±11

www.elsevier.com/locate/marpetgeo

Economic geology of offshore gas hydrate accumulations and provinces Alexei V. Milkov*, Roger Sassen Geochemical and Environmental Research Group (GERG), Texas A & M University, 833 Graham Road, Mail Stop 3149, College Station, TX 77845-668, USA Received 16 August 2001; received in revised form 5 November 2001; accepted 12 November 2001

Abstract The economic potential of well-studied offshore gas hydrate accumulations and provinces is assessed qualitatively based on consideration of geological, technological, and economic factors. Three types of gas hydrate accumulations are suggested. Structural accumulations occur where thermogenic, bacterial, or mixed gases are rapidly transported from the subsurface petroleum system to the gas hydrate stability zone along faults, mud volcanoes, and other structures (e.g. northwestern Gulf of Mexico, Hydrate Ridge, and Haakon Mosby mud volcano). These accumulations are generally characterized by high gas hydrate concentration in sediment, high resource density, high recovery factors, as well as low development and production costs. It is likely that structural accumulations provide marginal or economic gas hydrate reserves if they represent signi®cant volumes of hydrate-bound gas. Stratigraphic accumulations occur in relatively permeable sediments and form largely from bacterial methane generated in situ or slowly migrated from depth in the section (e.g. Blake Ridge, Gulf of Mexico minibasins). These accumulations are generally characterized by low gas hydrate concentration in sediments and low recovery factor, as well as high development and production costs. Stratigraphic accumulations mainly provide a subeconomic gas hydrate resource. However, in cases such as the Nankai Trough province, high gas hydrate concentration occurs in permeable sand layers and may represent a viable exploration and exploitation target. Less geological data are available on the combination gas hydrate accumulations controlled both by structures and stratigraphy. On the global scale, gas hydrate reserves are likely to represent only a small fraction of the gas hydrate resource because the largest volume of gas hydrate is in subeconomic stratigraphic accumulations. However, some concentrated gas hydrate accumulations may be exploited pro®tably, and those should be subjected to detailed quantitative economic analysis. q 2002 Elsevier Science Ltd. All rights reserved. Keywords: Gas hydrate; Energy resource; Hydrocarbon gases; Global

1. Introduction Gas hydrate is an energy mineral that occurs worldwide onshore in polar regions and offshore at water depths greater than ,200±600 m, depending on sea¯oor temperature and gas composition. Hydrate-bound gases are mainly methane but also include ethane through butanes hydrocarbons and non-hydrocarbon gases, such as CO2 and H2S. The hydrocarbon gases may provide more energy than all conventional subsurface fuel resources together (Kvenvolden, 1999). Knowledge of naturally occurring gas hydrate is increasing rapidly (Fig. 1). However, commercialization of gas hydrate remains unproven. Grauls (2001) predicts that gas production from onshore gas hydrate is not likely to commence prior to ,2010±2015 and from offshore gas hydrate before ,2030. Bil (2000) estimates the onset of * Corresponding author. Tel.: 11-979-862-2323x117; fax: 11-979-8622361. E-mail address: [email protected] (A.V. Milkov).

commercial gas hydrate recovery to be 2015 and 2060 for onshore and offshore gas hydrate, respectively. Geological, technological and economic issues are among factors that negatively affect the economic feasibility of gas hydrate recovery. Great uncertainty of the global gas hydrate resource (Lerche, 2000) and limited estimates of hydrate-bound gas in speci®c accumulations retard economic analysis of gas hydrate recovery. Offshore gas hydrate may contain from 0.2 £ 10 15 to 7600 £ 10 15 m 3 of gas (Fig. 2). Kvenvolden (1999) suggests that 21 £ 10 15 m 3 is a `consensus value' for the volume of offshore gas hydrate resource. This value is approximately 50 times greater than the world conventional gas endowment (0.436 £ 10 15 m 3, USGS World Energy Assessment Team, 2000, Fig. 2). However, Soloviev (2000) estimates that the global submarine gas hydrate resource is 0.2 £ 10 15 m 3, a value approximately half the world conventional gas endowment. Few studies have been carried out to estimate gas in place in individual gas hydrate accumulations. The volumes of

0264-8172/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved. PII: S 0264- 817 2( 01) 00047-2

2

A.V. Milkov, R. Sassen / Marine and Petroleum Geology 19 (2002) 1±11

(2) generalize geological, technological and economic factors that affect the feasibility of gas hydrate recovery; and (3) assess which regions are characterized by the highest probability of pro®table gas hydrate recovery in the future. This study systematizes and updates knowledge of gas hydrate as a future energy resource and provides a basis for effective gas hydrate exploration. 2. Gas hydrate accumulations and provinces

Fig. 1. Number of publications with words `gas hydrate(s)' in titles. Data from the major GeoRef w database.

hydrate-bound gas associated with the Buzdag mud volcano in the Caspian Sea (Ginsburg & Soloviev, 1998) and the Haakon Mosby mud volcano in the Norwegian Sea (Ginsburg et al., 1999) are similar, in the range 3±4 £ 10 8 m 3. The volume of gas resources in these accumulations is equivalent to that of small conventional gas ®elds. Larger gas hydrate accumulations may occur in fault-associated settings in the Gulf of Mexico (Milkov & Sassen, 2000, 2001a,b,c) and at the Cascadia margin (Suess et al., 1999, 2001), but no estimates are yet reported. Economic analysis of gas hydrate recovery is also inhibited by a lack of resource characterization (3D geological models of sediment permeability and gas hydrate concentration, resource density, recovery factors) by poor understanding of technological (recovery methods), and economic (development and production costs, infrastructure) factors. The economic signi®cance of gas hydrate was emphasized based on volumetrics alone (Collett, 1995; Kvenvolden, 1999). The present paper assesses offshore gas hydrate accumulations and provinces from the standpoint of economic geology. The objectives of this study are to (1) provide insight to well-documented gas hydrate localities;

Fig. 2. Global estimates of offshore gas hydrate resource (data from Kvenvolden, 1999 with additions) and estimates of the world conventional gas endowment (data from USGS World Energy Assessment Team, 2000) versus the year in which the estimate was made.

A gas hydrate accumulation is de®ned as a localized occurrence of gas hydrate in sediments related to a geological structure and/or stratigraphic trap. Several gas hydrate accumulations within a basin are considered to be a gas hydrate province. For example, gas hydrates associated with the Buzdag mud volcano in the South Caspian Sea (Ginsburg & Soloviev, 1998) make up the Buzdag gas hydrate accumulation. There are many gas hydrate accumulations in the South Caspian Sea (Diaconescu, Kieckhefer, & Knapp, 2001; Ginsburg & Soloviev, 1998), and they form a gas hydrate province. However, limited geological data make it dif®cult at present to distinguish between speci®c gas hydrate accumulations and gas hydrate provinces in many regions, such as Hydrate Ridge, Blake Ridge, and the Nankai Trough. Fluid (gas and water) migration into the gas hydrate stability zone (GHSZ) plays a critical role in the formation of a gas hydrate accumulation (Clennell, Hovland, Booth, Henry, & Winters, 1999; Ginsburg & Soloviev, 1998). Rapid gas transport is required to concentrate gas in permeable sediments where gas hydrate crystallizes. Water transport is usually thought to be less important because water is virtually omnipresent in sediments. However, water may be a limiting factor for gas hydrate crystallization at Hydrate Ridge (Suess et al., 2001) and in the Gulf of Mexico. The source of water for crystallization of gas hydrate in the Haakon Mosby mud volcano is thought to be from 2 to 3 km depth, rather than from in situ water (Ginsburg et al., 1999). Three types of gas hydrate accumulations are distinguished based on the mode of ¯uid migration and gas hydrate concentration within the GHSZ (Fig. 3). Structural accumulations occur where fault systems, mud volcanoes and other geological structures facilitate rapid ¯uid transport from depth into the GHSZ. In stratigraphic accumulations, gas hydrate crystallizes in relatively permeable strata from bacterial gas generated in situ or which is slowly supplied from great depth. These are end-members, and combination accumulations controlled both by structures and stratigraphy may occur. Later, we summarize typical characteristics of gas hydrate accumulations and compare their economic potential. 2.1. Structural accumulations Structural gas hydrate accumulations occur in advective

A.V. Milkov, R. Sassen / Marine and Petroleum Geology 19 (2002) 1±11

3

Fig. 3. Three types of gas hydrate accumulations (cross-sectional views). Arrows show ¯uid migration. Not to scale. See the text for details.

high ¯uid ¯ux settings (Xu & Ruppel, 1999; Fig. 3). Accumulations in the northwestern Gulf of Mexico (Brooks et al., 1986; MacDonald et al., 1994; Milkov & Sassen, 2000, 2001a; Sassen et al., 1999b, 2001), Hydrate Ridge (Hovland, Lysne, & Whiticar, 1995; Suess et al., 1999, 2001; Trehu, Torres, Moore, Suess, & Bohrmann, 1999), and the Haakon Mosby mud volcano (Bogdanov et al., 1999; Ginsburg et al., 1999) are well-studied structural gas hydrate accumulations. They occur at different water depths and in different tectonic settings (Fig. 4, Table 1). These accumulations have many common features. Gas

hydrate was sampled near the sea¯oor and was observed outcropping at the sea¯oor. Hydrocarbon gas vents from the sea¯oor to the water column, and chemosynthetic communities are common. In shallow sediments, gas hydrate largely occurs as plates, nodules, hailstone-type pellets, and massive vein-®llings in fracture porosity, and also ®lls matrix porosity. Gas hydrate concentration in sediments is relatively high because of rapid transport of gas from great depth along highly permeable fractured conduits. Both structure I and II gas hydrates crystallize from gas of thermogenic, bacterial, and mixed origin (Fig. 5). Methane

4

A.V. Milkov, R. Sassen / Marine and Petroleum Geology 19 (2002) 1±11

Fig. 4. Known (solid circles and squares) and inferred (open circles and squares) gas hydrate accumulations and provinces offshore (circles) and onshore (squares) (after Kvenvolden, 1999 with additions). The locations of the gas hydrate accumulations and provinces discussed in the text are indicated.

is the dominant gas in the gas hydrate lattice, but other hydrocarbons may also be signi®cant components in structure II gas hydrate. In proli®c petroleum provinces, such as the Gulf of Mexico, gas hydrate is often associated with oil. The subsurface 3D morphology of gas hydrate accumulations is thought to be controlled by the geometry of ¯uid conduits, ¯uid ¯ux rate, gas composition, and the temperature ®eld. Gas hydrate is localized around active faults and craters of mud volcanoes. Layers of gas hydrate within the GHSZ are suggested to be thick because of the high ¯uid ¯ux rate (Xu & Ruppel, 1999). The presence of thermogenic heavy hydrocarbon gases signi®cantly increases the thickness of the GHSZ (Milkov & Sassen, 2000). However, high temperature is often associated with ¯uid advection and may totally eliminate the GHSZ in the areas of highest ¯uid ¯ux (Ginsburg et al., 1999). Increased salinity of migrating pore water further retards gas hydrate stability (Sloan, 1998). Bottom simulating re¯ectors (BSRs) are not common or they are patchy and displaced in structural accumulations (Kastner, 2001; Milkov & Sassen, 2000). This results because the accumulations are vertically stacked, do not typically seal much gas below the gas hydrate layer, and the disturbed base of the GHSZ does not parallel the sea¯oor.

2.2. Stratigraphic accumulations Stratigraphic gas hydrate accumulations generally occur in advective low ¯uid ¯ux settings or diffusion-dominated settings (Xu & Ruppel, 1999; Fig. 3). These accumulations occur, for example, in relatively permeable strata of the Blake Ridge (Dickens, Paull, Wallace, & The ODP Leg 164 Scienti®c Party, 1997; Paull, Matsumoto, & Wallace, 1996; Paull, Matsumoto, Wallace, & Dillon, 2000), the Nankai Trough (Matsumoto, Takedomi, & Wassada, 2001; Max, 2000; Takahashi, Yonezawa, & Takedomi, 2001), and the Gulf of Mexico minibasins (Milkov & Sassen, 2001a; P¯aum, Brooks, Cox, Kennicutt, & Sheu, 1986) (Fig. 4, Table 1). Gas hydrate usually crystallizes well below the sea¯oor. It occurs mainly as small crystals in pore space, but nodules and plate crystals are also observed. Gas hydrate tends to be widely disseminated through the GHSZ, and low gas hydrate concentrations are commonly measured. However, there are exceptions, such as the Nankai Trough where gas hydrate occupies up to 82% of pores in thin but very permeable sand layers (Uchida, personal communication). Gas hydrate crystallizes mainly as structure I from bacterial methane (Fig. 5). Methane is believed to be generated in situ or is slowly supplied from deeper in the section.

f

e

d

c

b

a

High (fracture) High Average to high Low None Low

Low (matrix) Low High High Well developed Low

High (fracture) High Low to average Low None Average to high

Low (matrix) Low High High None Low

Low to high (matrix) Low to high High Average to high None Average to high

Up to 6 £ 1013 f Up to 18:4 £ 108

High (fracture) High Low Low Well developed High

2:8 £ 1013 12 £ 108

2±3 £ 1012 1 £ 108

3 £ 108 1:7 £ 108

Not reported Not reported

700±3500 32 000 50±500 (data from one well and BSR)

8±11 £ 1012 4±5 £ 108

1000±4000 26 000 58±620 (data from 18 wells and BSR)

Mainly from bacterial methane generated in situ or slowly supplied from depth below Up to 1±2 Average 2, up to 14 Average 10 (?), up to 30

615±2500 22 500 20±1500 (data from one well and modeling)

Nankai Trough e

700±1000 1250±1260 375 1.8 0±200 (data 0±160 (data from cores and from cores and modeling) BSR) From thermogenic, bacterial, and mixed gas rapidly migrated from depth below Average 20±30, up to 100 Up to 20±60 Up to 25

440±2500 23 000 0±1900 (data from cores and modeling)

Blake Ridge d

Northwestern Gulf of Mexico a

Haakon Mosby mud volcano c

Northwestern Gulf of Mexico a

Hydrate Ridge b

Stratigraphic accumulations/provinces

Structural accumulations/provinces

Data from Milkov and Sassen (2001a), P¯aum et al. (1986), and Sassen et al. (1999a,b). Data from Suess et al. (1999, 2001) and Trehu et al. (1999). Data from Bogdanov et al. (1999) and Ginsburg et al. (1999). Data from Dickens et al. (1997), Hollister and Ewing (1972), Paull et al. (1996, 2000), Sheridan and Gradstein (1983), and Keigwin et al. (1998). Data from Matsumoto et al. (2001), Takahashi et al. (2001), and Matsumoto (personal communication). The value is roughly calculated from the estimates of gas hydrate concentration around the Nankai Trough exploratory well and the areal extent of BSR distribution (Matsumoto, personal communication).

Water depth (m) Areal extent (km 2) Subsurface depth of gas hydrate occurrence (m) Gas hydrate origin Gas hydrate concentration (vol%) Resource (m 3) Average resource density (m 3/km 2) Permeability Recovery factor Development costs Production costs Infrastructure Economic potential

Characteristics

Table 1 Main geological, technological, and economic characteristics of gas hydrate accumulations and provinces A.V. Milkov, R. Sassen / Marine and Petroleum Geology 19 (2002) 1±11 5

6

A.V. Milkov, R. Sassen / Marine and Petroleum Geology 19 (2002) 1±11

Fig. 5. Relationship between mean d 13C (C1) and d D (C1) values for gas in gas hydrates sampled in the accumulations and provinces discussed in the text (data after Milkov and Sassen, 2001c; Paull et al., 2000; Lein, Vogt, Crane, Egorov, & Ivanov, 1999; Kastner et al., 1998; T. Uchida, personal communication). Only mean d 13C (C1) value is available for the Nankai Trough (shown by the dashed line). Fields of bacterial and thermogenic gas are depicted after Whiticar (1999).

The 3D morphology of gas hydrate accumulations is largely controlled by the geometry of relatively permeable strata and ¯uid transport into and within the GHSZ. Gas hydrate is localized in coarse-grained sediments (Ginsburg & Soloviev, 1998) that facilitate gas migration and provide space for gas hydrate nucleation (Clennell et al., 1999). Furthermore, gas hydrate appears to be concentrated in thin layers near the base of the GHSZ. The highest gas hydrate concentrations are likely to occur in low ¯ux advective settings (Xu & Ruppel, 1999). BSRs often occur at the base of stratigraphic gas hydrate accumulations because of their wide areal extent and the presence of free gas below the GHSZ (Hovland, 2000; Laherrere, 2000). 2.3. Combination accumulations In combination accumulations, gas hydrate occurs in relatively permeable strata, but gas is rapidly supplied from depth in the section along active faults or diapirs (Fig. 3). Diaconescu and Knapp (2000) and Diaconescu et al. (2001) report geophysical evidence for a combination gas hydrate accumulation in the Caspian Sea. A gas hydrate layer approximately 200 m thick is thought to be located at a subbottom depth of 300±350 m. Shallow faults and an active mud volcano occur near the gas hydrate accumulation

apparently serving as conduits for ¯uid ¯ow into the GHSZ. It is likely that gas hydrate crystallizes in relatively permeable sediments from migrated gas. Gas hydrate accumulations in the Hydrate Ridge, Blake Ridge and the Nankai Trough may also be a combination. More detailed geological and geophysical studies are needed to de®ne the full spectrum of factors that control gas hydrate occurrence in these geological settings.

3. Gas hydrate resource/reserve classi®cation Because gas hydrate is a mineral (Milkov, 2000; Sassen et al., 1999b), it is proposed that the USGS resource/reserve classi®cation scheme for minerals (US Bureau of Mines & the US Geological Survey, 1980) be applied to gas hydrate accumulations and provinces. The classi®cation is based on two criteria: (1) geological assurance of gas hydrate occurrence and (2) economic feasibility of gas hydrate recovery (Fig. 6). The arrangement of speci®c resource/reserve categories is modi®ed to better de®ne how gas hydrate accumulations and provinces may be classi®ed in one category or another when geological assurance and economic feasibility increase (Fig. 6).

A.V. Milkov, R. Sassen / Marine and Petroleum Geology 19 (2002) 1±11

Fig. 6. Resource/reserve classi®cation scheme for natural gas hydrates with examples.

3.1. Geological assurance The following categories are de®ned based on the increasing degree of geological assurance (Fig. 6). Gas hydrate resource is a concentration of gas hydrate in sediments in such form and amount that economic extraction of natural gas from it is currently or potentially feasible. Undiscovered (speculative and hypothetical) gas hydrate resources encompass gas hydrate accumulations, the existence of which is only postulated but not proved by direct or indirect methods. Speculative gas hydrate resources include undiscovered gas hydrate accumulations that may be present in geological settings where conditions are thought to be favorable for gas hydrate crystallization but no direct or indirect evidence for gas hydrate has been found so far. Any area characterized by pressure±temperature conditions favorable for gas hydrate crystallization and suf®cient amount of organic carbon in sediments provide speculative gas hydrate resources. Such areas occur mainly on continental slopes and rises (Ginsburg & Soloviev, 1998). Hypothetical gas hydrate resources include undiscovered gas hydrate accumulations that are similar to known gas hydrate accumulations and that may be reasonably expected to exist in similar geological features or provinces with analogous geological settings. Gas hydrate accumulations associated with deep-water mud volcanoes provide a good example of a hypothetical resource. These accumulations occur worldwide and may contain 10 10 ±10 12 m 3 of gas (Milkov, 2000). The estimated resource is hypothetical because it is an analogy between sampled hydrate-bearing mud volcanoes and other geologically similar features. Identi®ed (inferred, indicated, and measured) gas hydrate

7

resources are resources whose location, main characteristics, and quantity are known or estimated from speci®c geological, geophysical or geochemical evidence. Inferred gas hydrate resources include gas hydrate accumulations that are thought to occur at a speci®c location but have not been directly sampled by coring or drilling. Regions where geophysical (e.g. BSRs), geochemical (e.g. chlorinity anomalies), and geological (e.g. slumps) evidence for gas hydrate were found provide inferred gas hydrate resources. There are, so far, a total of 53 offshore areas worldwide where such gas hydrate accumulations are inferred (Fig. 4). Indicated gas hydrate resources include gas hydrate accumulations that have been sampled by coring, drilling, or by research submersible. The volume of hydrate-bound gas cannot be adequately estimated because of lack of detailed sampling. Indicated gas hydrate resources are present at 21 offshore localities worldwide (Fig. 4). At many sites, gas hydrate was recovered in only one or two drill holes or cores which limits accurate estimation of resources. Measured gas hydrate resources encompass gas hydrate accumulations and provinces in which the volume of hydrate-bound gas is de®ned by drill holes and cores, and the content of gas hydrate in sediments is estimated from detailed sampling. Few gas hydrate accumulations and provinces satisfy these conditions and provide measurable gas hydrate resources. Although the volume of hydratebound gas is estimated at most localities discussed in this paper (Table 1, Fig. 6), the 3D morphology of accumulations and provinces need to be better constrained in terms of gas hydrate concentration, among other factors. The northwestern Gulf of Mexico, Hydrate Ridge, Haakon Mosby mud volcano, Blake Ridge, and the Nankai Trough are characterized by the highest degree of geological assurance and may be subjected to economic analysis. 3.2. Economic feasibility The following categories are de®ned based on the increasing degree of economic feasibility of gas hydrate recovery (Fig. 6). Subeconomic gas hydrate resources are that part of identi®ed resources that do not meet the economic criteria of economic reserves and marginal reserves. Gas hydrate reserves (marginal and economic) are that part of identi®ed gas hydrate resources that may be economically extracted or produced at the time of determination. Marginal gas hydrate reserves include gas hydrate accumulations that border on being economically producible at the time of determination but would be producible if economic or technological factors change. Economic gas hydrate reserves include gas hydrate accumulations from which natural gas might be produced pro®tably under de®ned investment assumptions. Numerous geological, technological, and economic factors affect the economic feasibility of gas hydrate recovery. Although more detailed quantitative economic analysis is needed to

8

A.V. Milkov, R. Sassen / Marine and Petroleum Geology 19 (2002) 1±11

determine the reserve status of gas hydrate accumulations, the preliminary qualitative analysis below may be useful for directing further gas hydrate exploration. 3.2.1. Northwestern Gulf of Mexico Both structural and stratigraphic accumulations occur in the northwestern Gulf of Mexico (Milkov & Sassen, 2001a). Structural accumulations have high gas hydrate concentrations and signi®cant resource volumes. Average resource density over the Gulf of Mexico is not high (Table 1), but is suggested to be much greater in speci®c accumulations and intervals (Milkov & Sassen, 2001c). Gas hydrate occurs in shallow sediments with high fracture porosity and permeability that may result in high recovery. Development costs are suggested to be low because accumulations are located at relatively shallow water depth and are near the sea¯oor (Table 1). Thermal stimulation in combination with chemical inhibition (Sloan, 1998) may be used to recover gas hydrate from structural accumulations (Milkov & Sassen, 2001c). Horizontal drilling seems to be a viable technology in gas hydrate recovery. Production costs are suspected to be low because not much hot water or steam would be needed to decompose concentrated gas hydrate near equilibrium stability conditions. Finally, a well-developed petroleum infrastructure may decrease the cost of gas transportation. These favorable factors suggest that structural gas hydrate accumulations have high economic potential. It is probable that these accumulations provide marginal or economic reserves, and more detailed economic analysis is needed to certain if gas hydrate recovery is pro®table. In contrast, stratigraphic gas hydrate accumulations located in minibasins in the Gulf of Mexico are characterized by low (if any) economic potential. Gas hydrate concentration, resource, and resource density are insigni®cant (Table 1). The recovery factor is suggested to be low because of low depositional permeability. Gas hydrate accumulations are situated at great water depth and are likely to occur well below the sea¯oor that resulting in relatively high development costs. The gas hydrate resource is most likely to be subeconomic. 3.2.2. Hydrate Ridge The gas hydrate resource in this accumulation is not estimated. However, the signi®cant areal extent of Hydrate Ridge and the high gas hydrate concentration in sediments may imply that the volume of hydrate-bound gas is similar to reserves of a large or giant gas ®eld. High fracture permeability may result in a high average recovery factor. Development costs and production costs are suggested to be relatively low because gas hydrate occurs near the sea¯oor at relatively shallow water depth (Table 1). Economic potential of this accumulation may be high. However, no petroleum infrastructure exists at Hydrate Ridge, and signi®cant investment is required to transport hydrate-derived gas to markets. High concentration of H2S in the hydrate-bound gas (as much as 10%, Kastner,

Kvenvolden, & Lorenson, 1998) requires speci®c technologies to ensure safe recovery and transportation. These factors may decrease the economic feasibility of gas hydrate exploitation. Hydrate Ridge likely provides marginal reserves (Table 1, Fig. 6), and more detailed economic analysis is needed. 3.2.3. Haakon Mosby mud volcano The Haakon Mosby mud volcano is a well-studied gas hydrate accumulation. Twenty-seven cores were taken from the area ,1.8 km 2, and the sur®cial gas hydrate distribution is relatively well de®ned (Bogdanov et al., 1999; Ginsburg et al., 1999). Gas hydrate resource is insigni®cant (Table 1) because of the small areal extent of the accumulation and shallow GHSZ. Even though gas hydrate concentration is high in some sediment intervals, the volume of hydratebound gas is similar to the reserves of small conventional gas ®elds. This, in addition to the lack of nearby petroleum infrastructure and high development costs, implies that the gas hydrate resource at the Haakon Mosby mud volcano is subeconomic. 3.2.4. Blake Ridge The volume of hydrate-bound gas in this province is signi®cant (Table 1) and additional gas resources occur below the GHSZ as free gas. Based on volumetrics alone, Dickens et al. (1997) suggest that the Blake Ridge has enough methane to meet the 1996 US gas consumption rate for more than a hundred years. However, preliminary economic analyses suggest that gas hydrate there is not likely to be economic (Milkov & Sassen, 2001c). Gas hydrate is widely disseminated in sediments, and its concentration is low. The recovery factor is suggested to be insigni®cant because of low depositional permeability. Furthermore, the water depth is great and gas hydrate occurs at relatively great depth in the section (Table 1). Production of gas hydrate would be expensive. The depressurization method (Sloan, 1998) is unlikely to be used because insuf®cient free gas is trapped below the hydrate layer (Dickens et al., 1997; Laherrere, 2000). Thermal stimulation may require the injection of vast volumes of hot water or steam to decompose widely disseminated gas hydrate. Horizontal drilling does not seem to be a viable technology in gas hydrate recovery because of unfavorable ratio of water depth and subsurface depth of gas hydrate occurrence. Gas yield per well and per platform is suspected to be low. High development and production costs and the lack of existing petroleum infrastructure imply that gas hydrate recovery from the Blake Ridge area will not be costeffective. This province probably includes only subeconomic gas hydrate resources. Gas hydrate accumulations similar to the Blake Ridge were studied off Nigeria by Hovland, Gallagher, Clennell, and Lekvam (1997). Average gas hydrate concentration in stratigraphic accumulations above BSRs is suggested to be only 3 vol%. Hovland et al. (1997) concluded that there is

A.V. Milkov, R. Sassen / Marine and Petroleum Geology 19 (2002) 1±11

no commercial potential for recovery of gas hydrate because it is ®nely disseminated in sediments. 3.2.5. Nankai Trough This province has the highest gas hydrate resource and the most favorable resource density (Table 1). High gas hydrate concentrations occur in permeable sand layers, and the recovery factor is thought to be relatively high. Signi®cant water depth and subsurface depth of gas hydrate recovery imply that development costs may be high. Production costs are dif®cult to generalize because it is not clear, which gas hydrate recovery method would be appropriate for the Nankai Trough. Transportation of hydrate-derived gas would be expensive because of the lack of petroleum infrastructure there. However, the province is located near Japan where most natural gas is imported (Max, 2000), and its cost is much higher than, for example, in the U.S. This factor may signi®cantly increase the economic feasibility of gas hydrate recovery. The Nankai Trough may provide marginal or economic gas hydrate reserves, and more detailed economic analysis is needed. 4. Discussion On a global scale, structural gas hydrate accumulations that occur in high ¯ux advective settings are more likely to be economic than stratigraphic accumulations. High gas hydrate concentration, high resource density, and high recovery factor, and low development and production costs imply that structural accumulations may provide gas hydrate reserves. On the other hand, the small areal extent and shallow gas hydrate zone make many of them unattractive with regard to exploitation. For example, Hovland (2000) suggests that deep-water mud volcanoes represent a primary target for gas hydrate resource exploration. However, low gas hydrate resources in individual mud volcanoes (as it was suggested earlier in the Haakon Mosby mud volcano case study) imply that these gas hydrate accumulations are unlikely to be pro®tably exploited in the near future (Milkov, 2001). A continuous supply of gas from depth may result in the crystallization of fresh gas hydrate (Hovland, 2000) and thus, the gas hydrate resource at high-¯ux sites may be renewable. However, the kinetics of this process is unclear, and the economic signi®cance of the continuous replenishment of gas hydrates at such locations remains speculative. Although structural gas hydrate accumulations occur worldwide, most of them tend to be located on convergent margins (Fig. 4), where ongoing tectonic activity facilitates rapid upward and lateral ¯uid migration. Kastner (2001) estimates that 60±65% of the global gas hydrate resource is situated in convergent margins. In stratigraphic accumulations occurring in diffusiondominated and low ¯ux advective settings, gas hydrate is

9

often too disseminated to represent an economically viable exploration and exploitation target. Low gas hydrate concentration and a low recovery factor are characteristic of these accumulations. On the other hand, there are exceptions like the Nankai Trough province, where high gas hydrate concentrations occur in high-permeability sand reservoirs. Exploration for such accumulations may be viable. Many gas hydrate accumulations inferred to occur based on geophysical, geochemical, and geological evidence (Fig. 4) are likely to be stratigraphic. Moreover, many known gas hydrate accumulations shown in Fig. 4 are associated with relatively small geological structures, such as mud volcanoes (Milkov, 2000). This observation suggests that the global gas hydrate reserve represents only a small portion of the global gas hydrate resource. Most gas hydrate provinces and accumulations thus provide subeconomic gas hydrate resource, and a signi®cant increase in gas price is required to move this resource to the category of reserves. Gas hydrate is most likely to be pro®tably recovered in countries, where most natural gas is imported. For example, Japan has a very limited hydrocarbon resources, and imports ,97% of annually consumed gas (Max, 2000). Relatively high gas prices increase economic feasibility of gas hydrate recovery. In addition, the stability of the economy of Japan signi®cantly depends on the availability of its own hydrocarbon resources. It is suggested that the recovery of gas hydrate from the Nankai Trough may commence ,2010 (Max, 2000). The U.S. imported only 15.8% of total gas consumption in 1999 (Energy Information Administration, 2000), and it is relatively independent from foreign gas suppliers. However, projected growth of gas demand, importation, and price (Energy Information Administration, 2000) may in the future increase the economic feasibility of gas hydrate recovery from the U.S. Economic Exclusive Zone. Safety during gas hydrate recovery is a pivotal issue. Seabed instability may be a result of gas hydrate dissociation because physical properties of sediment change, water and gas content in sediments increases, and shear strength decreases (Campbell, 1991; Hovland & Gudmestad, 2001; Milkov, Sassen, Novikova, & Mikhailov, 2000). The geohazard issue must be carefully addressed in structural gas hydrate accumulations located near the sea¯oor in underconsolodated sediments, such as those in the Gulf of Mexico and in the Hydrate Ridge. Moreover, the association of unique biologically valuable chemosynthetic communities with gas hydrate has important environmental implications (Sassen et al., 1999a). As the Hydrate Ridge and the Blake Ridge case studies suggest, the degree of infrastructure development is an important factor that signi®cantly affects the economic feasibility of gas hydrate recovery. Gas hydrate provinces and accumulations located within producing petroleum provinces (e.g. northwestern Gulf of Mexico, Caspian Sea, Niger Delta) may be more cost-effective to exploit because

10

A.V. Milkov, R. Sassen / Marine and Petroleum Geology 19 (2002) 1±11

less investments are required to transport hydrate-derived gas. On the other hand, it is possible to construct a gas hydrate recovery and transportation system that involves a specially equipped vessel (Oil & Gas Journal, 1999). In this case, no great modi®cation of existing infrastructure is required to exploit the most remote gas hydrate accumulations. 5. Conclusions The preliminary qualitative economic analysis described here has many assumptions. However, the economic potential of well-studied gas hydrate accumulations and provinces varies widely. Gas hydrate recovery may be economical only where concentrated gas hydrate deposits occur in relatively permeable sediments. From the standpoint of economic geology, attractive gas hydrate resources are thought to occur mostly in structural accumulations (e.g. northwestern Gulf of Mexico, Hydrate Ridge). However, the low volumes of hydrate-bound gas associated with some structural accumulations (e.g. Haakon Mosby mud volcano) signi®cantly decrease the economic feasibility of gas hydrate recovery. Stratigraphic accumulations do not provide much gas hydrate reserves. Gas hydrate occurs in low concentration and is widely disseminated in sediments (e.g. Blake Ridge, Gulf of Mexico minibasins). However, in areas where permeable coarse-grained sediments are present, and a signi®cant volume of methane is available in the GHSZ (e.g. the Nankai Trough), gas hydrate recovery may be pro®table. Globally, the gas hydrate reserve is likely to represent only a small portion of the gas hydrate resource. This observation results because many gas hydrate accumulations are stratigraphic, occur in diffusion-dominated and low ¯ux advective settings, and have small gas hydrate concentrations in sediments. In addition, many structural gas hydrate accumulations provide only insigni®cant subeconomic resources. It is suggested that initial gas hydrate exploration should concentrate on relatively large structural accumulations. The main preliminary ®nding of this study is that gas hydrate may be pro®tably recovered from some accumulations. Detailed quantitative economic analysis is needed to prove this conclusion using the approach of economic geology. Acknowledgements R. Matsumoto and T. Uchida are sincerely acknowledged for providing as yet unpublished data from the Nankai Trough. M. Hovland is thanked for constructive review. The research is supported by the Applied Gas Hydrate Research Program at Texas A & M University. Additional support was provided by the American Association of Petro-

leum Geologists (Grants-in-Aid 2000 Program), the Gulf Coast Association of Geological Societies (Student Grants in 2000 and 2001), and the Geological Society of America (Student Research Grant in 2001). References Bil, K. J. (2000). Economic perspective of methane from hydrate. In M. D. Max, Natural gas hydrate in oceanic and permafrost environments (pp. 349±360). Dordrecht: Kluwer Academic Publishers. Bogdanov, Yu. A., Sagalevich, A. M., Vogt, P. R., Mienert, J., Sundvor, E., Crane, K., Lein, A. Yu., Egorov, A. V., Peresypkin, V. I., Cherkashev, G. A., Gebruk, A. V., Ginsburg, G. D., & Voitov, D. V. (1999). The Haakon Mosby mud volcano in the Norwegian sea: Results of multidisciplinary studies with manned submersibles. Oceanology, 39, 374± 380. Brooks, J. M., Cox, H. B., Bryant, W. R., Kennicutt II, M. C., Mann, R. G., & McDonald, T. J. (1986). Association of gas hydrates and oil seepage in the Gulf of Mexico. Organic Geochemistry, 10, 221±234. Campbell, K. J. (1991). Deepwater geohazards: An engineering challenge. Offshore, 51, 46±51, see also p. 60. Clennell, M. B., Hovland, M., Booth, J. S., Henry, P., & Winters, W. J. (1999). Formation of natural gas hydrates in marine sediments. 1. Conceptual model of gas hydrate growth conditioned by host sediment properties. Journal of Geophysical Research, 104 (B10), 22985±23003. Collett, T. S. (1995). Gas hydrate resources of the US. In D. L. Gauter, G. L. Doulton, K. I. Takahishi & K. L. Varnes (Eds.), National assessment of US oil and gas resources on CD-ROM USGS Digital Data Series 30. Diaconescu, C. C., & Knapp, J. H. (2000). Buried gas hydrates in the deepwater of the South Caspian Sea, Azerbaijan: Implications for geohazards. Energy Exploration and Exploitation, 18, 385±400. Diaconescu, C. C., Kieckhefer, R. M., & Knapp, J. H. (2001). Geophysical evidence for gas hydrates in the deep water of the South Caspian Basin, Azerbaijan. Marine and Petroleum Geology, 18, 209±221. Dickens, G. R., Paull, C. K., Wallace, P. The ODP Leg 164 Scienti®c Party & (1997). Direct measurement of in situ methane quantities in a large gas-hydrate reservoir. Nature, 385, 426±428. Energy Information Administration (2000). The annual energy outlook 2001. Report DOE/EIA-0383. http://www.eia.doe.gov/oiaf/aeo/index. html (accessed in February 2001). Ginsburg, G. D., & Soloviev, V. A. (1998). Submarine gas hydrates, St Petersburg: VNIIOkeangeologia, 215 pp. Ginsburg, G. D., Milkov, A. V., Soloviev, V. A., Egorov, A. V., Cherkashev, G. A., Vogt, P. R., Crane, K., Lorenson, T. D., & Khutorskoy, M. D. (1999). Gas hydrate accumulation at the Haakon Mosby mud volcano. Geo-Marine Letters, 19, 57±67. Grauls, D. (2001). Gas hydrates: Importance and applications in petroleum exploration. Marine and Petroleum Geology, 18, 519±523. Hollister, C. D., & Ewing, J. I. (1972). Initial Reports of the Deep Sea Drilling Project, Vol. XI. Washington, DC: US Government Printing Of®ce. Hovland, M. (2000). Are there commercial deposits of marine hydrates in ocean sediments? Energy Exploration and Exploitation, 18, 339±347. Hovland, M., & Gudmestad, O. T. (2001). Potential in¯uence of gas hydrates on seabed installations. In C. K. Paull & W. P. Dillon, Natural gas hydrates: Occurrence, distribution, and dynamics. AGU Monograph Series, Vol. 24. (pp. 307±315). Hovland, M., Lysne, D., Whiticar, & M. J. (1995). Gas hydrate and sediment gas composition, Hole 892A. In B. Carson, G. K. Westbrook, R. J. Musgrave, & E. Suess (Eds.), Proceedings of the Ocean drilling Program, Scienti®c Results, Vol. 146 (pp. 151±161). Hovland, M., Gallagher, J. W., Clennel, M. B., & Lekvam, K. (1997). Gas hydrate and free gas volumes in marine sediments: Example from the Niger Delta front. Marine and Petroleum Geology, 14, 245±255. Kastner, M. (2001). Gas hydrates in convergent margins: Formation,

A.V. Milkov, R. Sassen / Marine and Petroleum Geology 19 (2002) 1±11 occurrence, geochemistry, and global signi®cance. In C. K. Paull & W. P. Dillon, Natural gas hydrates: Occurrence, distribution, and dynamics. AGU Monograph Series, Vol. 24 (pp. 67±86). Kastner, M., Kvenvolden, K. A., & Lorenson, T. D. (1998). Chemistry, isotopic composition, and origin of methane±hydrogen sul®de hydrate at the Cascadia subduction zone. Earth and Planetary Science Letters, 156, 173±183. Keigwin, L. D., Rio, D., & Acton, G. D. (1998). Proceedings on ODP, Initial Reports, Vol. 172. College Station, TX (Ocean Drilling Program). Kvenvolden, K. A. (1999). Potential effects of gas hydrate on human welfare. Proceedings of the National Academy of Science, USA, 96, 3420±3426. Laherrere, J. (2000). Oceanic hydrates: More questions than answers. Energy Exploration and Exploitation, 18, 349±384. Lein, A., Vogt, P., Crane, K., Egorov, A., & Ivanov, M. (1999). Chemical and isotopic evidence for the nature of the ¯uid in CH4-containing sediments of the Haakon Mosby mud volcano. Geo-Marine Letters, 19, 76±83. Lerche, I. (2000). Estimates of worldwide gas hydrate resource. Energy Exploration and Exploitation, 18, 329±337. MacDonald, I. R., Guinasso, N. L., Sassen, R., Brooks, J. M., Lee, L., & Scott, K. T. (1994). Gas hydrates that breaches the sea-¯oor on the continental slope of the Gulf of Mexico. Geology, 22, 699±702. Matsumoto, R., Takedomi, Y., & Wassada, H. (2001). Exploration of marine gas hydrates in Nankai Trough, offshore Central Japan. AAPG Annual Convention, Of®cial Program, 10, A128. Max, M. D. (2000). Hydrate as a future energy resource for Japan. In M. D. Max, Natural gas hydrate in oceanic and permafrost environments (pp. 225±238). Dordrecht: Kluwer Academic Publishers. Milkov, A. V. (2000). Worldwide distribution of submarine mud volcanoes and associated gas hydrates. Marine Geology, 167, 29±42. Milkov, A. V. (2001). Gas hydrate: A common mineral in deep-water mud volcanoes. Causes, Processes, and Effects of Subsurface Sediment Mobilization on Reservoir to Regional Scale, Abstract Book, Gent 2001 (p. 57). Milkov, A. V., & Sassen, R. (2000). Thickness of the gas hydrate stability zone, Gulf of Mexico continental slope. Marine and Petroleum Geology, 17, 981±991. Milkov, A. V., & Sassen, R. (2001a). Estimate of gas hydrate resource, northwestern Gulf of Mexico continental slope. Marine Geology, 179, 71±83. Milkov, A. V., & Sassen, R. (2001b). Structurally focused gas hydrate a future northwest US Gulf resource. Offshore, 61 (9), 92±94. Milkov, A. V., & Sassen, R. (2001c). Economic geology of the Gulf of Mexico and the Blake Ridge gas hydrate provinces. Transactions Gulf Coast Association of Geological Societies, 51, 219±228. Milkov, A. V., Sassen, R., Novikova, I., & Mikhailov, E. (2000). Gas hydrates at minimum stability water depth in the Gulf of Mexico: Signi®cance to geohazard assessment. Transactions Gulf Coast Association of Geological Societies, 50, 217±224. Oil & Gas Journal (1999). Syntroleum unveils hydrate recovery process, 97 (44). Paull, C. K., Matsumoto, R., & Wallace, P. J. (1996). Proceedings on ODP, Initial Reports, Vol. 164, College Station, TX (Ocean Drilling Program). Paull, C. K., Matsumoto, R., Wallace, P. J., & Dillon, W. P. (2000).

11

Proceedings on ODP, Scienti®c Results, Vol. 164, College Station, TX (Ocean Drilling Program). P¯aum, R. C., Brooks, J. M., Cox, H. B., Kennicutt, M. C. II, & Sheu, D.-D. (1986). Molecular and isotopic analysis of core gases and gas hydrates, DSDP Leg 96. In A. H. Bouma, J. M. Coleman, & A. W. Meyer (Eds.), Initial Reports of the Deep Sea Drilling Project, 96. Washington, DC: US Government Printing Of®ce (pp. 781±784). Sassen, R., Joye, S., Sweet, S. T., DeFreitas, D. A., Milkov, A. V., & MacDonald, I. R. (1999a). Thermogenic gas hydrates and hydrocarbon gases in complex chemosynthetic communities, Gulf of Mexico continental slope. Organic Geochemistry, 30, 485±497. Sassen, R., Sweet, S. T., Milkov, A. V., DeFreitas, D. A., Salata, G. G., & McDade, E. C. (1999b). Geology and geochemistry of gas hydrates, central Gulf of Mexico continental slope. Transactions Gulf Coast Association of Geological Societies, 49, 462±468. Sassen, R., Losh, S., Cathles, L., Roberts, H., Whelan, J. K., Milkov, A. V., Sweet, S. T., & DeFreitas, D. A. (2001). Massive vein-®lling gas hydrate: Relation to ongoing gas migration from the deep subsurface Gulf of Mexico. Marine and Petroleum Geology, 18, 551±560. Sheridan, R. E., & Gradstein, F. M. (1983). Initial Reports of the Deep Sea Drilling Project, Vol. 76. Washington: US Government Printing Of®ce. Soloviev, V. A. (2000). Global estimation of gas content in submarine gas hydrate accumulations. Sixth International Conference on Gas in Marine Sediments, St Petersburg (pp. 123±125). Sloan, E. D. (1998). Clathrate hydrates of natural gases, (2nd ed). New York: Marcel Dekker, 705 pp. Suess, E., Torres, M. E., Bohrmann, G., Collier, R. W., Greinert, J., Linke, P., Rehder, G., Trehu, A., Wallmann, K., Winckler, G., & Zuleger, E. (1999). Gas hydrate destabilization: Enhanced dewatering, benthic material turnover and large methane plumes at the Cascadia convergent margin. Earth and Planetary Science Letters, 170, 1±15. Suess, E., Torres, M. E., Bohrmann, G., Collier, R. W., Rickert, D., Gold®nger, C., Linke, P., Heuser, A., Sahling, H., Heeschen, K., Jung, C., Nakamura, K., Greinert, J., Pfannkuche, O., Trehu, A., Klinkhammer, G., Whiticar, M. J., Eisenhauer, A., Treichert, B., & Elvert, M. (2001). Sea ¯oor methane hydrates at Hydrate Ridge, Cascadia Margin. In C. K. Paull & W. P. Dillon, Natural gas hydrates: Occurrence, distribution, and dynamics. AGU Monograph Series, Vol. 24. (pp. 87±98). Takahashi, H., Yonezawa, T., & Takedomi, Y. (2001). Exploration for natural hydrate in Nankai-Trough wells offshore Japan. Offshore Technology Conference Proceedings, Vol. 1 (OTC 13040). Trehu, A. M., Torres, M. E., Moore, G. F., Suess, E., & Bohrmann, G. (1999). Temporal and spatial evolution of a gas hydrate-bearing accretionary ridge on the Oregon continental margin. Geology, 27, 939±942. US Bureau of Mines & the US Geological Survey (1980). Principles of a resource/reserve classi®cation for minerals. US Geological Survey Circular 831. USGS World Energy Assessment Team (2000). US Geological Survey world petroleum assessment 2000Ðdescription and results. USGS Digital Data Series DDS-60. Whiticar, M. J. (1999). Carbon and hydrogen isotope systematics of bacterial formation and oxidation of methane. Chemical Geology, 161, 291±314. Xu, W., & Ruppel, C. (1999). Predicting the occurrence, distribution, and evolution of methane gas hydrate in porous marine sediments. Journal of Geophysical Research, 104, 5081±5096.